Cyclopropylindole-based synthetic cannabinoids, including UR-144, 5F-UR-144 (XLR-11), and XLR-12, represent a structurally atypical subgroup of novel psychoactive substances that remain insufficiently characterized in terms of their toxicological properties, despite increasing relevance in clinical and forensic toxicology. The lack of systematic data on their toxicity constitutes a critical gap, particularly given their unique molecular architecture, which may influence biological interactions and toxicological outcomes. In this study, we performed a comprehensive preclinical toxicological characterization with a focus on molecular aspects of toxicity, using an integrative set of qualitative and quantitative in silico approaches. Qualitative analyses (STopTox, ADMETlab 3.0, admetSAR 3.0, ProTox 3.0) were applied to evaluate acute toxicity (oral, dermal, inhalation), genotoxicity (Ames test), and organ-specific toxicity, including effects on the cardiovascular, renal, pulmonary, and hematological systems, alongside identification of structural toxicophores. Quantitative methods (VEGA QSAR, Percepta) provided estimates of acute toxicity across species and exposure routes, as well as cardiotoxic potential related to hERG channel inhibition. The results indicate that specific molecular features play a key role in shaping the toxicological profiles of the investigated compounds. In particular, the cyclopropyl moiety was consistently associated with increased cardiotoxic potential, while fluorinated substituents (5-fluoropentyl in XLR-11 and trifluorobutyl in XLR-12) modulated predicted toxicity, including acute toxicity and genotoxicity-related endpoints. All compounds demonstrated a high probability of pulmonary toxicity (> 85%) and moderate to high nephrotoxic potential (~ 70%), suggesting potential relevance for inhalation exposure and renal risk observed in clinical settings. Importantly, the integration of toxicophore analysis with quantitative toxicity parameters enabled the identification of structure–toxicity relationships at the molecular level, providing a basis for mechanistic interpretation and supporting the development of testable hypotheses for future preclinical studies. This work contributes to the toxicological characterization of emerging psychoactive substances and highlights the importance of molecular features in determining their potential health risks.